Re: [PATCH 06/11] mm: page_vma_mapped_walk(): crossing page table boundary

[Hugh Dickins]

On Thu, 10 Jun 2021, Kirill A. Shutemov wrote:
> On Wed, Jun 09, 2021 at 11:44:10PM -0700, Hugh Dickins wrote:
> > page_vma_mapped_walk() cleanup: adjust the test for crossing page table
> > boundary - I believe pvmw->address is always page-aligned, but nothing
> > else here assumed that;
> 
> Maybe we should just get it aligned instead? (PMD_SIZE - PAGE_SIZE) is not
> most obvious mask calculation.

Would you prefer it with another line of comment after the
/* Did we cross page table boundary? */

Maybe,
/* Is address always page-aligned? No need to assume that. */

I just don't see the point in forcing alignment when the test is good
(and don't know for sure whether address is always aligned there or not).

I prefer to leave it as is, just letting the commit message document it.

> 
> > and remember to reset pvmw->pte to NULL after
> > unmapping the page table, though I never saw any bug from that.
> 
> Okay, it's fair enough.

Thanks,
Hugh